Fractal technique for abstracting high-resolution information from GPR section

We present a method for abstracting the high-resolution information from Ground Penetrating Radar (GPR for short) detecting section. In this method, the random fractal technique based on fractal Brown movement (fBm for short) has been introduced into GPR data processing. It has been demonstrated that GPR time series meet certain requirements of the fBm characteristics. We could get the fractal parameters of GPR time series by establishing the corresponding GPR fBm model, and compute the fractal parameters. We easily reconstruct the GPR fractal dimension section without any crease using the fractal parameters by adopting successive random additions algorithm (SRA). In this paper the slope interface model testing has been performed. Applications of the algorithm to synthetic and field GPR data produce GPR section of underground objects. Our results have shown that the fractal reconstruct of GPR section with a properly fractal parameters can abstract meaningful high-resolution information from GPR section. Compared with the conventional signal processing method, it has certain distinct advantages.